Prenatal exposure to fine particulate matter PM2.5 and small for gestational age: a Bayesian model for area-based data in Milan

Air pollution is a known risk factor for adverse birth outcomes, including Small for Gestational Age (SGA) births. This study examines the association between fine particulate matter (PM2.5) exposure and SGA births in Milan, Italy, considering spatial dependencies and socioeconomic factors. We applied a Bayesian hierarchical spatial model with a binomial regression framework to birth data aggregated at a 500 m × 500 m grid level. A Conditional Autoregressive (CAR) prior captured spatial correlations. Covariates included maternal age, Deprivation Index, Normalized Difference Vegetation Index (NDVI), surface temperature, and Road Coverage. Parameter estimation was performed using Markov Chain Monte Carlo (MCMC) methods. Among 7635 eligible births in 2016, 8.5% were SGA. A 10 µg/m3 increase in PM2.5 was associated with a 15% increase in SGA odds (OR: 1.153, IQR: 0.853–1.556). The D eprivation Index also showed a strong positive association (OR: 1.075, IQR: 1.028–1.125). NDVI exhibited a weak positive association, potentially reflecting socioeconomic disparities. Maternal age, temperature, and Road Coverage were not significantly associated with SGA. PM2.5 exposure and socioeconomic deprivation are linked to higher SGA risk in Milan. The spatial correlation highlights localized risk factors. Targeted policies to reduce air pollution and address social inequalities are needed to improve perinatal outcomes.